dataset
active
dataset:dinov3

DINOv3

Main vision model whose activations are decomposed.

Neighborhood — ranked by edge-count